Abstract Details


Poster 24: Prediction of Thermal Properties of Organic Peroxides Using QSPR Models

Vinca Prana1, 2, Patricia Rotureau1, Guillaume Fayet1, Carlo Adamo2, 3
1INERIS, Parc Technologique Alata, BP2, 60550 Verneuil-en-Halatte, France
2Laboratoire d’Electrochimie, Chimie des Interfaces et Modélisation pour l’Energie, CNRS UMR-7575, Chimie ParisTech, 11 rue P. et M. Curie, 75231 Paris, France
3Institut Universitaire de France, 103 Boulevard Saint Michel, F-75005 Paris, France
The new EU regulation REACH (Registration, Evaluation, Authorisation and Restriction of Chemicals) aiming to reinforce the control of risks from chemicals in Europe is entered into force in 2007. Before 2018, in order to allow the use on the market of every substance produced or imported for more than one ton per year in Europe, the evaluation of toxicological, eco-toxicological and physico-chemical properties is required. Taking into account the number of substances and properties as well as other factors (timing, economic costs, feasibility at the R&D level and risks for the manipulator) the measurement of all the data is not realistic. Thus, the development of alternative predictive methods for the evaluation of the properties of substances was recommended in the framework of REACH.
In this context, the French PREDIMOL [1] (molecular modelling prediction of physico-chemical properties of products) project funded by ANR has started in November 2010 for 3 years. Its objective is to demonstrate that molecular modelling is a credible alternative method to experiment to obtain missing physico-chemical data for REACH regulation and also to gain recognition of them by regulatory European instances.
In particular, we focused on Quantitative Structure-Property Relationship (QSPR) models which have been recommended in REACH framework. These models were developed according to the five OCDE principles drawn up for the validation of QSPR models:
1. A defined endpoint (including experimental protocol);
2. An unambiguous algorithm;
3. A defined domain of applicability;
4. Appropriate measures of goodness-of fit, robustness and predictive power;
5. A mechanistic interpretation, when it’s possible.
An original approach associating the QSPR method to quantum chemical calculations was used with the aim to answer more easily to the 5th principle. More than 300 molecular descriptors (constitutional, topological, geometrical, quantum chemical) were calculated using CodessaPro software from calculated molecular structures, optimized with the Density Functional Theory (DFT) in Gaussian09 package. This approach has already been used successfully by our team for the prediction of the heat of decomposition of nitroaromatic compounds [2,3] and the prediction of the impact sensitivity of nitroaliphatic compounds [4].
In this study some predictive and validated models for the heat and temperature of decomposition of organic peroxides were developed based on a database of 38 organic peroxides completely developed in the PREDIMOL framework (acquisition of reference experimental data using Differential Scanning Calorimetry apparatus). A particular attention was paid to guarantee that data were measured in homogenous experimental conditions. The influence of the concentration in organic peroxides was also observed for the prediction of the heat of decomposition. Models were developed using two methods: PLS (partial least square) and MLR (multi-linear regression). This latter easier to apply gave QSPR models with better performances in terms of fitting, robustness and predictivity for these both properties. Moreover some descriptors of these models are linked to the peroxide bond which is related to the mechanism of decomposition of organic peroxides starting by the cleavage of the peroxide bond. To our knowledge, these models are also the first which were rigorously validated by following OECD principles for regulatory acceptability of QSPRs, so that their application in the REACH framework is possible.

[1] www.ineris.fr/predimol
[2] J. Mol. Model., 17, 2011, 2443-253
[3] Chem. Phys. Lett., 467, 2009, 407-411
[4] J. Hazard. Mater., 235– 236, 2012, 169– 177

Return to Programme